All DAC chips sample music in a particular way, regardless if the music is recorded at 16bit/44.1kHz (normal 'redbook' CD standard) or 24bit/192kHz (the current typical HD standard). It requires a filter after the conversion from digital data to analog sound to cut off all the signal around and above 20kHz so the noise generated by the DAC will be filtered out. Thus there is no signal above the cut off point and it shouldn't matter much to you as a listener.
However, there is a certain advantage for a headphone to have extended frequency response. One, the further the FR extends, the flatter it tends to be at the treble region. Two, you could get better harmonic on the upper end and a richer tone. However, this is assuming the goal of the headphone is to be as flat and extended as possible, and most headphones are not. Headphone is different from speaker as you listen to them very closely, so you don't have any kind of attenuation naturally caused by the air and environment - this makes headphone to sound overly bright, thus a little roll off at the top might be preferred. This comes to another issue - spec are often misleading, as manufacturer tends to list what ever range they like without tolerance or graph. So it could be -3dB at 20kHz or -30dB at 20dB, and that kind of difference is between one you can hear, and one you can't hear (but still measurable). So take spec with a grain of salt.
